1 Basic details about project

1.1 Abbreviations

  • IC = intermittent claudication
  • RP = rest pain
  • TL = tissue loss
  • TASCII/TASC-II/TASC = Trans-Atlantic Inter-Society Consensus-II
  • Bal = Balloon angioplasty
  • DEB = Drug-eluting balloon
  • St = Stent
  • Sup = Supera
  • VIA = VIABAHN
  • Fail = Failed procedure
  • CFA = common femoral artery
  • P3 = P3 segment of popliteal artery
  • Lysis = intra-arterial thrombolysis
  • Tech succ = technical success
  • Complex no. = number of complexity markers

1.2 Inclusion and exclusion criteria

  • Indication criteria are:
    • Endovascular intervention to femoro-popliteal (fem-pop) segment
    • Intermittent claudication, rest pain and tissue loss
  • Exclusion criteria are:
    • Femoro-popliteal bypass graft on ipsilateral leg
    • Acute limb ischaemia or other indication not listed above
    • Previous intervention to same limb is NOT an exclusion criterion

1.3 Period covered

  • Procedures from 2016-01-11 to 2020-11-25 are included here

2 Patient information

2.1 Patients per year

  • 579 are included
  • They are spread across the years as demonstrated:

2.2 Age per year

  • Mean age of patients is 70.6 years
  • Median age of patients is 72 years
  • Age is demonstrated in this boxplot:

  • Age variation by year is demonstrated in this boxplot:

  • To check if age is evenly distributed across the years, they can be plotted to identify outliers:

  • Data points 164, 396 and 431 are outliers
  • Point 164: 71 years old (2017)
  • Point 396: 72 years old (2019)

  • Point 431: 68 years old (2019)

  • Welch’s one-way test performed instead of ANOVA due to inhomogeneity of variances, looking for a significant age difference between years:

## 
##  One-way analysis of means (not assuming equal variances)
## 
## data:  patients$Age and as.factor(patients$Year)
## F = 0.68578, num df = 4.00, denom df = 277.99, p-value = 0.6023

2.3 Age per indication

  • Average age by indication for intervention is as shown here:
Indication Mean age Median age
IC 64.0 65
RP 70.4 72
TL 72.0 73
  • These data are demonstrated in this boxplot:

  • To check if these data are evenly distributed, they can be plotted to identify outliers:

  • Data points 396, 431 and 488 are outliers
  • Point 396: 72 years old (TL)
  • Point 431: 68 years old (IC)

  • Point 164: 73 years old (TL)

  • Welch’s one-way test performed instead of ANOVA due to inhomogeneity of variances:

## 
##  One-way analysis of means (not assuming equal variances)
## 
## data:  patients$Age and patients$Indication
## F = 25.446, num df = 2.00, denom df = 151.53, p-value = 2.963e-10
  • Pairwise t-test with no assumption of equal variances to look for where differences in means are, given that Welch’s test was signficant:
## 
##  Pairwise comparisons using t tests with non-pooled SD 
## 
## data:  patients$Age and patients$Indication 
## 
##    IC      RP     
## RP 0.00036 -      
## TL 1.2e-10 0.26689
## 
## P value adjustment method: BH

2.4 Age per gender

  • Average age by gender for intervention is as shown here:
Gender Mean age Median age
Man 72.5 75
Woman 69.4 71
  • These data are demonstrated in this boxplot:

  • t-test test to look for differences between mean ages:
## 
##  Welch Two Sample t-test
## 
## data:  patients$Age by patients$Gender
## t = 3.2047, df = 427.3, p-value = 0.001453
## alternative hypothesis: true difference in means is not equal to 0
## 95 percent confidence interval:
##  1.235113 5.153242
## sample estimates:
##   mean in group Man mean in group Woman 
##            72.54911            69.35493

3 Limb information

3.1 Limbs per year

  • 632 are included
  • They are spread across the years as demonstrated:

4 Procedure information

4.1 Procedures per year

  • 719 are included
  • On average, roughly 148 procedures are performed each year
  • Or 12 per month
  • They are spread across the years as demonstrated:

4.2 TASC classification of lesions

  • These are from the 719 procedures

5 Indication information

  • Number of procedures performed for each indication are shown here:

  • Number of procedures each year for each indication are shown here:

5.1 Intermittent claudication

  • On average, roughly 22 procedures are performed each year for IC
  • Or 1.8 per month

5.1.1 By TASC

5.1.2 By TASC each year

5.2 Rest pain

  • On average, roughly 20 procedures are performed each year for RP
  • Or 1.6 per month

5.2.1 By TASC

5.2.2 By TASC each year

5.3 Tissue loss

  • On average, roughly 108 procedures are performed each year for TL
  • Or 8.9 per month

5.3.1 By TASC

5.3.2 By TASC each year

6 Technical success

6.1 Overall technical success

  • 90.7% of all fem-pop procedures are technically successful

6.2 Technical success each year

  • Kruskal-Wallis test to look for difference in success between years:
## 
##  Kruskal-Wallis rank sum test
## 
## data:  procedures$Successful by as.factor(procedures$Year)
## Kruskal-Wallis chi-squared = 1.3559, df = 4, p-value = 0.8518
  • Wilcoxon rank sum test to look for where differences in means are:
  • (can be ignored if p >= 0.05 for Kruskal-Wallis test)
## 
##  Pairwise comparisons using Wilcoxon rank sum test 
## 
## data:  procedures$Success and as.factor(procedures$Year) 
## 
##      2016 2017 2018 2019
## 2017 0.92 -    -    -   
## 2018 0.92 0.92 -    -   
## 2019 0.92 0.92 0.97 -   
## 2020 0.92 0.92 0.92 0.92
## 
## P value adjustment method: BH

6.3 Technical success by indication

  • Kruskal-Wallis test to look for difference in success between indications:
## 
##  Kruskal-Wallis rank sum test
## 
## data:  procedures$Successful by procedures$Indication
## Kruskal-Wallis chi-squared = 0.55031, df = 2, p-value = 0.7595
  • Wilcoxon rank sum test to look for where differences in means are:
  • (can be ignored if p >= 0.05 for Kruskal-Wallis test)
## 
##  Pairwise comparisons using Wilcoxon rank sum test 
## 
## data:  procedures$Success and procedures$Indication 
## 
##    IC   RP  
## RP 0.71 -   
## TL 0.71 0.71
## 
## P value adjustment method: BH

6.4 Technical success by TASC-II

  • Kruskal-Wallis test to look for difference in success between indications:
## 
##  Kruskal-Wallis rank sum test
## 
## data:  procedures$Successful by procedures$TASCII
## Kruskal-Wallis chi-squared = 49.491, df = 3, p-value = 1.026e-10
  • Wilcoxon rank sum test to look for where differences in means are:
  • (can be ignored if p >= 0.05 for Kruskal-Wallis test)
## 
##  Pairwise comparisons using Wilcoxon rank sum test 
## 
## data:  procedures$Success and procedures$TASCII 
## 
##   A       B       C     
## B 0.9602  -       -     
## C 0.0388  0.0129  -     
## D 1.6e-07 1.1e-10 0.0034
## 
## P value adjustment method: BH

7 Treatment method

7.1 All procedures

  • Overall, including failed procedures, 79% of treatments leave nothing behind

  • More specific devices:

7.2 Device by year

  • More specific devices:

7.3 Device by indication

  • The bar chart is separated into three with different y-axes to allow easy comparison of proportions

  • More specific devices:

7.4 Device by TASC-II

  • More specific devices:

8 Procedural complexity

  • Five indicators of increased procedural complexity are recorded, with a binary "yes"/"no" option being available, ie. the success of the added procedural step is not recorded:
  1. ‘Iliac’ - iliac angioplasty or stent in the same procedure
  2. ‘CFA’ - concurrent common femoral endarterectomy
  3. ‘P3’ - P3/below knee segment of the popliteal artery was treated
  4. ‘Crural’ - concurrent infra-popliteal angioplasty
  5. ‘Lysis’ - acute intra-procedural thrombus requiring intra-arterial thrombolysis

8.1 Number and percentage of individual complexity markers

Yes/No Iliac CFA P3 Crural Lysis
No 696.0 672.0 586.0 533.0 706.0
Yes 23.0 47.0 133.0 186.0 13.0
% Yes 3.2 6.5 18.5 25.9 1.8

8.2 Number of procedures with total number of complexity markers

Number of complexity markers Number Percentage
0 420 58.4
1 203 28.2
2 89 12.4
3 7 1.0

8.2.1 Procedural success by complexity

  • NB success is defined as successfully treating the fem-pop segment

  • Kruskal-Wallis test to look for difference in success between complexity levels:
## 
##  Kruskal-Wallis rank sum test
## 
## data:  procedures$Successful by procedures$multiComp
## Kruskal-Wallis chi-squared = 6.5881, df = 3, p-value = 0.08625
  • Wilcoxon rank sum test to look for where differences in means are:
  • (can be ignored if p >= 0.05 for Kruskal-Wallis test)
## 
##  Pairwise comparisons using Wilcoxon rank sum test 
## 
## data:  procedures$Success and procedures$multiComp 
## 
##   0     1     2    
## 1 0.099 -     -    
## 2 0.533 0.533 -    
## 3 0.533 0.533 0.533
## 
## P value adjustment method: BH

9 Survival

  • Death dates were last checked 17/04/2020

  • Survival statistics are based on the 579 patients in the data
  • Kaplan-Meier survival estimator is used for survival curves

  • Log-rank test is used to sompare survival estimates between groups

9.1 Overall

  • Kaplan-Meier survival for all patients undergoing fem-pop intervention is plotted below:

9.2 By year

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  patients and Year 
## 
##      2016 2017 2018 2019
## 2017 0.94 -    -    -   
## 2018 0.94 0.94 -    -   
## 2019 0.98 0.94 0.98 -   
## 2020 0.94 0.94 0.94 0.94
## 
## P value adjustment method: BH

9.3 By TASC-II

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  patients and TASCII 
## 
##   A       B       C      
## B 0.00031 -       -      
## C 1.4e-06 0.03079 -      
## D 0.00011 0.14135 0.89372
## 
## P value adjustment method: BH

9.4 By Gender

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  patients and Gender 
## 
##       Man 
## Woman 0.72
## 
## P value adjustment method: BH

9.5 By indication

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  patients and Indication 
## 
##    IC      RP   
## RP 1.6e-05 -    
## TL 1.1e-09 0.014
## 
## P value adjustment method: BH

9.6 By technical success

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  patients and Successful 
## 
##     No    
## Yes 0.0046
## 
## P value adjustment method: BH

9.7 By treatment method

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  patients and Dev 
## 
##      Bal   St   
## St   0.753 -    
## Fail 0.013 0.058
## 
## P value adjustment method: BH

9.8 By complexity

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  patients and multiComp 
## 
##   0    1    2   
## 1 0.73 -    -   
## 2 0.73 0.73 -   
## 3 0.73 0.73 0.73
## 
## P value adjustment method: BH

9.9 Tissue loss only

9.9.1 By TASC-II

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  tlPat and TASCII 
## 
##   A      B      C     
## B 0.0199 -      -     
## C 0.0023 0.1099 -     
## D 0.0155 0.3987 0.8733
## 
## P value adjustment method: BH

9.9.2 By Gender

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  tlPat and Gender 
## 
##       Man 
## Woman 0.53
## 
## P value adjustment method: BH

9.9.3 By technical success

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  tlPat and Successful 
## 
##     No   
## Yes 0.004
## 
## P value adjustment method: BH

9.9.4 By treatment method

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  tlPat and Dev 
## 
##      Bal   St   
## St   0.895 -    
## Fail 0.012 0.048
## 
## P value adjustment method: BH

9.9.5 By complexity

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  tlPat and multiComp 
## 
##   0    1    2   
## 1 0.55 -    -   
## 2 0.55 0.55 -   
## 3 0.55 0.55 0.55
## 
## P value adjustment method: BH

9.10 Thirty day survival by indication

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  patients and Indication 
## 
##    IC    RP   
## RP 1.000 -    
## TL 0.088 0.088
## 
## P value adjustment method: BH

10 Amputation

  • Amputation dates were last checked 22/06/2020

  • Survival statistics are based on the 632 limbs in the data
  • Kaplan-Meier survival estimator is used for survival curves

  • Log-rank test is used to compare amputation survival estimates between groups

10.1 Overall

  • Kaplan-Meier survival for all limbs undergoing fem-pop intervention is plotted below:

10.2 By year

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  limbs and Year 
## 
##      2016  2017  2018  2019 
## 2017 0.884 -     -     -    
## 2018 0.055 0.066 -     -    
## 2019 0.673 0.483 0.033 -    
## 2020 0.055 0.055 0.682 0.033
## 
## P value adjustment method: BH

10.3 By TASC-II

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  limbs and TASCII 
## 
##   A       B       C      
## B 0.02242 -       -      
## C 0.06906 0.95016 -      
## D 0.00033 0.03848 0.06906
## 
## P value adjustment method: BH

10.4 By Gender

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  patients and Gender 
## 
##       Man   
## Woman 0.0053
## 
## P value adjustment method: BH

10.5 By indication

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  limbs and Indication 
## 
##    IC      RP     
## RP 0.00014 -      
## TL 1.5e-06 0.10276
## 
## P value adjustment method: BH

10.6 By technical success

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  limbs and Successful 
## 
##     No  
## Yes 0.14
## 
## P value adjustment method: BH

10.7 By treatment method

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  limbs and Dev 
## 
##      Bal  St  
## St   0.43 -   
## Fail 0.29 0.29
## 
## P value adjustment method: BH

10.8 By complexity

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  limbs and multiComp 
## 
##   0     1     2    
## 1 0.729 -     -    
## 2 0.086 0.178 -    
## 3 0.357 0.357 0.729
## 
## P value adjustment method: BH

10.9 Tissue loss only

10.9.1 By TASC-II

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  tlLimb and TASCII 
## 
##   A      B      C     
## B 0.0665 -      -     
## C 0.2443 0.5126 -     
## D 0.0063 0.0983 0.0983
## 
## P value adjustment method: BH

10.9.2 By Gender

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  tlLimb and Gender 
## 
##       Man   
## Woman 0.0018
## 
## P value adjustment method: BH

10.9.3 By technical success

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  tlLimb and Successful 
## 
##     No   
## Yes 0.072
## 
## P value adjustment method: BH

10.9.4 By treatment method

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  tlLimb and Dev 
## 
##      Bal  St  
## St   0.57 -   
## Fail 0.15 0.15
## 
## P value adjustment method: BH

10.9.5 By complexity

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  tlLimb and multiComp 
## 
##   0    1    2   
## 1 0.28 -    -   
## 2 0.47 0.28 -   
## 3 0.47 0.28 0.47
## 
## P value adjustment method: BH

10.10 Thirty day amputation survival by indication

## Warning: Vectorized input to `element_text()` is not officially supported.
## Results may be unexpected or may change in future versions of ggplot2.

  • Log rank test to look for significant survival difference between groups:
## 
##  Pairwise comparisons using Log-Rank test 
## 
## data:  limbs and Indication 
## 
##    IC    RP   
## RP 0.006 -    
## TL 0.006 0.719
## 
## P value adjustment method: BH

Home